Presentation 2000/11/2
Associative Memory Using Neurons Having Refractory Period
Mikio ODA, Hiromi MIYAJIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In autocorrelation associative memory using a neural netwoek, monotone activation function is used in order to guarantee the network convergence. However, the network has disadvantage of having poor memory capacity and spurious memory while it has advantage of the simplicity of its structure of the memory. In the paper, autocorrelation associative memory using a neural network with neurons having refractory period is proposed, and the proposed memory is evaluated by computer simulation compared with conventional models. It shows that the proposed network has much more memory capacity than hopfield network on the condition of guaranteeing the network convergence. In addition, the network is also shown to has more memory capacity than a conventional improved model, nonmonotone activation function model in the case that the network does not guarantee to converge.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Associative Memory / Autocorrelation / Refractory Period / Monotone Function
Paper # CAS 2000-64,CST 2000-19
Date of Issue

Conference Information
Committee CST
Conference Date 2000/11/2(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Concurrent System Technology (CST)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Associative Memory Using Neurons Having Refractory Period
Sub Title (in English)
Keyword(1) Associative Memory
Keyword(2) Autocorrelation
Keyword(3) Refractory Period
Keyword(4) Monotone Function
1st Author's Name Mikio ODA
1st Author's Affiliation Kurume National College of Technology()
2nd Author's Name Hiromi MIYAJIMA
2nd Author's Affiliation Faculty of Engineering, Kagoshima University
Date 2000/11/2
Paper # CAS 2000-64,CST 2000-19
Volume (vol) vol.100
Number (no) 417
Page pp.pp.-
#Pages 8
Date of Issue